Nobody Wants to Be the Trust That Applies the Deduction

The structural flaw of contractors marking their own homework

Why facilities management contract governance fails in PFI and the NHS, and why the bill for it is about to arrive.

Every PFI contract in the NHS estate contains a payment mechanism. Each one sets out, in exhaustive schedule-level detail, what constitutes a service failure, how long the contractor has to put it right, what happens if they do not, and precisely what it costs them when they fail. The contractual power to enforce performance has been sitting in those schedules since financial close, available to any authority willing to use it.

The National Audit Office nevertheless found weak contract enforcement in around half of the trusts it sampled, and reported that authorities were frequently disinclined to impose penalties at all. It went further than that, noting a tendency among trusts in financial difficulty to ease the performance requirements placed on project companies in exchange for a reduced unitary charge, which amounts to waiving the contractual transfer of risk in order to help resolve a deficit. Its conclusion was uncomfortably direct, warning that contractors will seek to pass risk back to the taxpayer wherever contracts are not enforced.

The explanation offered for this within the sector is almost always the same, and it is couched in the language of partnership. Deductions sour the relationship, collaboration achieves more than confrontation, and an informal resolution is said to deliver a better outcome than a formal one. That explanation is doing a considerable amount of work, and it deserves rather more scrutiny than it usually receives.

The capability underneath the culture

Forty-eight per cent of trusts did not meet the standard for contract management staffing, and twelve per cent were unable to commit even a single day a week to monitoring their contracts. Those figures describe the constraint sitting underneath the cultural explanation, and they are difficult to reconcile with the idea that non-enforcement is a considered strategic choice.

It is very difficult to enforce a performance regime that you do not have the people to measure, and it is effectively impossible to defend a deduction that you cannot evidence. An authority which knows it cannot evidence a deduction has a strong incentive to describe its restraint as a relationship decision rather than a resourcing failure, because the first of those is a strategy and the second is an audit finding.

That is the uncomfortable core of facilities management contract governance across much of the public sector. In the majority of cases the decision not to enforce is not taken from a position of strength, but because the client-side function lacks the specialist expertise, the independent monitoring data, or simply the hours in the week to carry an argument through to a conclusion. The contractual remedy exists in full, while the organisational capacity required to operate it does not.

Once that becomes established practice it tends to compound. A payment mechanism that is never applied gradually stops functioning as a control and becomes a formality, and both parties quickly learn what the operative rules really are. The specification continues to say one thing while the operating culture says another, and the distance between the two widens with every year the contract runs.

The structural flaw of contractors marking their own homework

The enforcement problem sits on top of a design problem, and that design problem is self-monitoring. Most PFI performance regimes require the project company to monitor its own performance and report the results to the authority, which leaves the authority overseeing a return it did not generate, drawn from a helpdesk system it does not control, and categorised against failure definitions that the reporting party is itself interpreting. The incentives created by that arrangement have been obvious for twenty-five years.

The clearest illustration remains the litigation between Compass Group and Mid Essex Hospital Services NHS Trust, in which the contractor omitted pages of a health and safety inspection report that identified areas requiring immediate improvement. The case is usually cited for what it established about good faith and the exercise of contractual discretion, but it is at least as instructive for what it reveals about the information asymmetry built into the model, because the authority simply did not know what it had not been told.

Where authorities have invested in genuinely independent monitoring, the results tend to speak for themselves. Monitoring evidence has proved decisive in demonstrating a project company’s failure to rectify issues logged to the helpdesk, and has supported both structured remediation programmes and the application of deductions. Evidence changes the nature of the conversation entirely, and without it the authority is left negotiating on the contractor’s version of events.

The scale of what accumulates in the absence of that evidence is visible in the market. One NHS trust has tendered PFI estates rectification work covering more than forty thousand issues identified across a series of surveys, which is not a maintenance backlog in any ordinary sense but the accumulated residue of a performance regime that was never really operating.

Why this stops being an abstraction in 2030

For most of the PFI era the cost of weak enforcement was both deferred and diffuse, which is a large part of why it was tolerated for so long, but that cost is now becoming concentrated and, more importantly, dated.

Around one hundred and forty PFI contracts expire before 2030, and the National Audit Office guidance is that handback planning should begin seven years ahead of expiry, which means that for a substantial share of that cohort the planning window has already closed. In March 2026 the National Infrastructure and Service Transformation Authority published new contract management guidance for PFI, expecting formal Senior Responsible Owner governance of expiry as a programme and including an explicit capability diagnostic for the intelligent client function. The direction of travel is unambiguous, in that the centre now treats client-side capability as the principal variable determining outcomes at expiry.

The reason is straightforward enough. At handback, every lifecycle obligation that was never enforced, every rectification that was logged and quietly absorbed, and every condition standard that was allowed to drift becomes a survey finding and then a negotiation. By that stage the authority leverage has very largely evaporated, because deductions that were not applied cannot be applied retrospectively, the service continuity risk sits with the authority rather than the contractor, and the contractor remaining exposure is both limited and time-bound.

This is landing on an estate with no capacity to absorb it. NHS maintenance backlog liability now stands at around £15.9bn, with the high-risk element at £3.5bn following a rise of twenty-eight per cent in a single year, and forty-three of the two hundred and fifteen trusts in England carry backlog bills of £100m or more. For organisations in that position a handback dispute over asset condition is not an administrative matter but a capital problem they have no capital available to solve.

The same design, being bought again

It would be comfortable to file all of this as a legacy PFI issue that will retire itself alongside the contracts that created it, but that would be a mistake. RM6378, the Crown Commercial Service framework for facilities management and security services now used by most public sector buyers, carries self-monitoring performance requirements of its own. The model in which the supplier measures and reports its own performance while a thinly resourced client function reviews the output is not being phased out at all, but is being re-procured on new paper.

The question for any authority letting facilities management services through that route is whether it is building the client-side function to go alongside it, or whether it is simply repeating, on a shorter term, precisely the arrangement whose consequences the PFI estate is now counting.

What actually changes the outcome

The remedies are unglamorous, which is part of the reason they are so easily deferred. Resourcing the intelligent client function properly, and treating it as an investment rather than an overhead, is the first of them. The NISTA capability diagnostic is a reasonable place to begin, if only because it produces a defensible internal case for the headcount, and a contract management function costing a few hundred thousand pounds a year is trivial when set against a handback dispute on a major acute site.

The second is to stop relying solely on self-reported performance data. Independent verification does not need to be exhaustive in order to be effective, but it does need to be credible and unpredictable enough to change reporting behaviour, and sample-based assurance is usually sufficient to achieve exactly that.

The third is to apply the mechanism at least occasionally, because a payment mechanism that has never once been used carries no deterrent value whatsoever. Applying it consistently to well-evidenced failures is not an act of aggression towards a supplier, it is the maintenance of a control that the organisation is otherwise paying for without receiving any benefit from.

The fourth is to separate the relationship from the remedy. The best-run contracts in the sector manage to apply deductions and sustain constructive supplier relationships at the same time, and those two things are not in tension with one another. Treating them as though they were is precisely how authorities talk themselves out of using the only leverage available to them.

The fifth, for any authority approaching expiry, is to start the programme immediately if it has not already begun. Seven years represents the minimum realistic runway required to identify condition problems, agree remediation, complete the works, secure replacement arrangements, manage workforce transition and resolve disputes, and anything materially less than that is not a programme but a scramble.

None of this requires a change in the law, a renegotiation, or an intervention from policymakers. The contractual authority already exists within the schedules, fully drafted and long since signed. What has been missing is the capability and the institutional willingness to use it, and for around one hundred and forty contracts there is now a finite and shrinking period in which using it will still make any difference at all.

Where independent support changes the equation

Most estates teams reading this will recognise the problem without needing it explained to them, and will also recognise that the binding constraint is rarely willingness. The obstacles are practical ones, being a shortage of specialist capacity, an absence of defensible evidence, and a wholly reasonable reluctance to put an operational relationship under strain when the same supplier is relied upon to keep theatres running and wards clean the following morning. Independent advisory support addresses each of those obstacles directly, and it does so without requiring the trust to create permanent headcount it has no realistic prospect of funding.

The most immediate benefit is the separation of the challenge from the relationship. When performance failures are raised by an independent adviser working to a defined brief, the conversation becomes a technical one about evidence and contractual definitions rather than a personal one between people who have to work alongside each other every day. The trust retains the relationship while the adviser carries the friction, and that separation is very often the single factor that allows a properly evidenced deduction to be applied at all.